* Support importing .docx as new documents
* Add html file support, build types and interface for easily adding file types to importer
* fix: Upload embedded images in docx to storage
* refactor: Bulk of logic to command
* refactor: Do all importing on server, so we're not splitting logic for import into two places
* test: Add documentImporter tests
Co-authored-by: Lance Whatley <whatl3y@gmail.com>
* fix: Server error if attempting to load an unknown attachment
* fix: Migration should cascade delete to document attachments
* fix: Delete S3 attachments along with documents
* save images as private and serve via signed url from images.info api
* download private images to directory on export
* fix lint errors
* private s3 default, AWS.s3 module level scope, default s3 url expiry
* combine regex to one, and only replace when there are matches
* fix lint
* code not needed anymore, remove
* updates after pulling master
* revert the uploadToS3FromUrl url return
* use model gettr to compact code, rename to attachments api
* basic checking of document read permission to allow attachment viewing
* fix: Continue to upload avatars as public
fix: Allow redirect for non-private attachments
* add support for publicly shared documents
* catch errors which crash the app during zip export and user creation
* add tests
* enable AWS signature v4 for s3
* switch to use factories to build models for testing
* add isDocker flag for local serving of attachment redirect url
* fix redirect tests
Co-authored-by: Tom Moor <tom.moor@gmail.com>
* intial commiy
* cleaning code
* added makeCredential to s3.js and removed extra module dependecy
* lint fixes
* minor fix
* minor fixes
* changed encoding type from string to any
* added new env var to env.sample
The max size should be an integer rather than a string (environment variables
are strings). More strict s3 implementations reject the upload because they do
not consider the policy valid if a string is used rather than an integer.
